Latest Polling Data in CD 8 Congressional Race

I do so love an interesting poll. Patty Weiss recently released the main thrust of her latest set of polls showing her with a statistically significant lead in the in the Democratic Primary. Patty came in at 23% strong support (32% overall), Gabby Giffords comes in at 16% strong support (22% overall), Jeff Latas and Alex Rodriguez vie for third place with 7 and 6% respectively, and Francine Shacter brings up the rear. The big winner in the poll is, of course, Mr. Undecided with a whooping 32%, though nobody seems committed to his candidacy. These numbers aren’t what’s interesting about this poll, however.

Of course, we all understand the limitations of polls. Heh. Little joke. Far too few understand the limits of polling. They are only as good as the accuracy of the demographic framing (the universe of people from which a sample is chosen). And they are only a snapshot in time, subsequent events can change the outcome greatly. And the questions one asks are almost as important to the answers one gets as people’s real attitudes. And the confidence intervals, if too large, can wreak havok with the reliability of the data.

Those disclaimers made, let’s look at this new poll. It is a Democratic registered only poll. In fact, it is a 4 by 4 Democrat universe of 14,000. That means that only the people who have voted in all four of the most recent four elections – the supervoters, if you will – are sampled. That makes the poll representative of the core of the Democratic Party in CD 8. This universe pretty fairly represents the expected Democratic Primary electorate in a non-Presidential year, when adjusted for further demographics such as age, sex, geographic location, and ethnicity.

Conducted by Lake Research Partners over a 4 days period, Patty’s poll sampled 400 interviewees for a 16-18 minute verified ID phone interview. Incomplete interviews were dumped. This smallish size sample results in a fairly high confidence interval of 4.9%. That means that there is a 95% chance that the real numbers in the universe population lie within + or – 4.9% of stated values. With Patty and Gabby’s numbers within 10% of each other, that means they could, in fact, be tied; though it is vastly more likely that Patty does have an appreciable lead over Gabby of anywhere from 1-10%.

The most important thing that this poll indicates is that this race isn’t over by any means. That 32% of Mr. Undecided could swamp any candidate, or carry any candidate over the finish line. Conceivably support for either Patty or Gabby could stagnate at current levels, and Latas, Rodriguez, or Shacter, if able to persuade Mr. Undecided’s voters, could carry the election. But in order for that to happen, the least informed, least committed, and most easily persuadable would have to break overwhelmingly for a candidate who lacks significant name recognition, and hasn’t the money for major media buys to increase their exposure. Possible, but short of a free-media bonanza, not very likely.

What is most likely is that the great majority of those remaining Undecideds will break for the candidate they’ve heard of, or seen an ad for. This is where Patty Weiss’ name ID of some 80% translates into a major capital asset. Patty doesn’t need to try nearly as hard fashioning an image and getting her name known, she already has both. Some may say that Patty’s image as a trusted newscaster doesn’t translate to electoral credibility. I’m not so sure: this isn’t exactly the best year to be seen as a professional politician. I think that the idea of a citizen representative doing service to his or her community by holding political office will be very appealing to most voters. For this reason, Gabby’s overwhelming fundraising advantage may be consumed in part just getting her name out there and framing her image.

It is hard to predict how the remaining Undecided voters will break without having seen any of the media of Patty and Gabby. I can and will predict that Gabby is going to have to spend much more heavily for every voter. Running from behind, she’s going to have to convert those voters at a greater rate to overtake Patty. When paired with Patty’s established reputation in the wider community (word of mouth from friends and relatives of all those Undecided Democrats) Patty’s head start indicated by this poll is a formidable advantage, even with less money. Unless Gabby can split the remaining Undecideds 2 to 1 in her favor, She’s going to be hard pressed to find the votes to win.

In addition, the supporters of Latas and Rodriguez tend to indicate their second choice is Patty or Undecided, while much fewer roll-off directly to Gabby. Should either, or both, of these candidates pull up their tent-stakes prior to the primary and endorse another candidate, or just release their voters, Patty would tend to be the beneficiary. Gabby can undoubtedly capture a lot of these voters, but again, she is looking at spending time (which may not be available) and resources recruiting voters that the poll indicates Patty will capture fairly easily. In fact, if either Latas or Rodriquez threw their support behind Patty, the primary race may well be effectively over. Gabby would then face an almost 20% deficit and would have to convert Undecideds at a blistering 3-1 rate. Patty would also be perilously close to gaining an absolute majority.

More than anything decisive in the numbers, Patty’s poll reveals some fairly thorny structural biases in public attitudes that make gaining momentum relatively more expensive and difficult for Gabby than for Patty. This race can be won by anyone, but even the most partisan activist will admit that victory by Patty or Gabby looks most likely. And between the two of them, Gabby will have to come from behind, spend at a greater rate for each new voter, and hope that every break breaks her way, not Patty’s. I will be looking under toadstools for other polls out there that might tell a different story, of course.